airbyte
ActiveDescription
Airbyte is an open-source ELT data integration platform designed for AI agents and the modern data stack, connecting APIs, databases, and files to warehouses and AI applications.
Key Features
- 600+ data connectors covering APIs, databases, warehouses, and lakes
- Designed for AI agents with Airbyte Agents and a dedicated Agent SDK
- No-code Connector Builder to create custom data sources in minutes
- Low-code CDK using YAML to define data sync logic
- Multiple deployment forms: open-source self-hosted, Airbyte Cloud, Enterprise Kubernetes
- LLM tool integration compatible with pydantic-ai, LangChain, OpenAI Agents, and FastMCP
Use Cases
💡 Connect AI agents to SaaS and enterprise data for real-time CRM, ticket, and database access
💡 Build enterprise ELT pipelines that consolidate distributed data into a warehouse
💡 Use the Agent SDK to wrap connectors as LLM-callable tools
💡 Replicate and sync data across systems with 600+ connectors
Strengths & Limitations
✅ Strengths
- • Actively maintained, recent updates
- • High community interest (21.9k stars)
⚠️ Limitations
- • High issue backlog (2.3k open issues)
- • No clear open-source license
Categories
Quick Start
# Install the Agent SDK
uv pip install airbyte-agent-sdk
# Create an LLM tool set
from airbyte_agent_sdk import AirbyteAgent
agent = AirbyteAgent(sources=["hubspot", "postgres"])
tools = agent.as_tools() # Returns LangChain / pydantic-ai compatible tools
# Use inside an agent
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(model="gpt-4").bind_tools(tools)
llm.invoke("List HubSpot contacts created last week")